Practical aspects of Deep Learning
Submit your assignment
Due DateMay 31, 2:59 AM EDT
Attempts3 every 8 hours
Receive grade
To Pass80% or higher
Grade
100%
We keep your highest score
Due May 31, 2:59 AM EDT
If you have 10,000,000 examples, how would you split the train/dev/test set?
The dev and test set should:
If your Neural Network model seems to have high bias, what of the following would be promising things to try? (Check all that apply.)
You are working on an automated check-out kiosk for a supermarket, and are building a classifier for apples, bananas and oranges. Suppose your classifier obtains a training set error of 0.5%, and a dev set error of 7%. Which of the following are promising things to try to improve your classifier? (Check all that apply.)
What is weight decay?
What happens when you increase the regularization hyperparameter lambda?
With the inverted dropout technique, at test time:
Increasing the parameter keep_prob from (say) 0.5 to 0.6 will likely cause the following: (Check the two that apply)
Which of these techniques are useful for reducing variance (reducing overfitting)? (Check all that apply.)
Why do we normalize the inputs x?